Lessons · 5
- How AI Gets Things WrongHallucination, confident tone, training cutoffs, and bias in training data.
- A Practical Verification FrameworkTier 1 light review, Tier 2 spot-check claims, Tier 3 verify before use.
- Data Privacy and What Not to Put in a PromptM365 Copilot vs public tools; PII, confidential, client, and legal content.
- Responsible Use in High-Stakes ContextsClient-facing, leadership, HR, attribution, review, and transparency.
- Building Good AI Judgment Over TimeOver-trust vs under-trust; habits; MillerKnoll data and review-before-send.